Re: 4NF is Where It Is At! [WAS Re: 1:1 relationships]

From: Jan Hidders <hidders_at_REMOVE.THIS.win.tue.nl>
Date: 26 Jan 2001 17:37:02 GMT
Message-ID: <94scju$3b5$1_at_news.tue.nl>


Barry wrote:
>
> And this is where I've concluded 4th Normal Form is actually the most
> important of all. At the risk of being flamed for overly-simplifying it,
> 4NF asks: do all non-key attributes have the *same* relationship to the
> key? If not, then split the non-key attributes apart.

I do not really know how to say this but I think you are over-simplifying. :-) The problem is that you are talking about non-key attributes where the real 4NF talks about *all* attributes. So your normal form is stronger than 3NF but weaker than 4NF and different from BCNF. So I suggest that we call it the BJNF and define it as follows:

  A relation is in BJNF if

  1. all multi-valued dependencies are functional dependencies, and
  2. it is in 3NF.

I conjecture that -- just as 3NF -- you can always reach BJNF while staying dependency preserving.

With the rest of what you wrote I fully agree.

-- 
  Jan Hidders
Received on Fri Jan 26 2001 - 18:37:02 CET

Original text of this message